__highs2half2

Applicability

Product	Supported
Atlas 350 Accelerator Card	√
Atlas A3 training product/Atlas A3 inference product	x
Atlas A2 training product/Atlas A2 inference product	x
Atlas 200I/500 A2 inference product	x
Atlas inference product AI Core	x
Atlas inference product Vector Core	x
Atlas training product	x

Function Usage

Extracts the high 16 bits from each of the two half2 inputs, fills them into half2, and returns the result.

Prototype

inline half2 __highs2half2(const half2 x, const half2 y)

Parameters

**Table 1** Parameters
Parameter	Input/Output	Description
x	Input	Source operand.
y	Input	Source operand.

Returns

Result returned after the high 16 bits are extracted from each of the two half2 inputs and filled into half2

Restrictions

For SIMT programming, this API is not supported.

Header Files to Be Included

To use this API, the simt_api/asc_fp16.h header file must be included.

#include "simt_api/asc_fp16.h"

Examples

For SIMD and SIMT programming:

// Use small vectors to improve data movement efficiency.
__simt_vf__ __launch_bounds__(1024) inline void simt_highs2half2(__gm__ half2* input1, __gm__ half2* input2, __gm__ half2* output, uint32_t input_total_length)
{
    uint32_t idx = blockIdx.x * blockDim.x + threadIdx.x;
// Each thread processes one piece of half2 data, that is, two pieces of half data. Therefore, threads with idx greater than or equal to input_total_length/2 do not process data.
    if (idx > input_total_length / 2) {
        return;
    }
    output[idx] = __highs2half2(input1[idx], input2[idx]);
}
__global__ __vector__ void cast_kernel(__gm__ half* input1,  __gm__ half* input2, __gm__ half* output, uint32_t input_total_length)
{
    asc_vf_call<simt_highs2half2>(dim3(1024), (__gm__ half2*)input1, (__gm__ half2*)input2, (__gm__ half2*)output, input_total_length);
}

Parent topic: Type Conversion